Recognizing the People Problem in AI and Big Data (The Designerly Leap Trilogy Part II)

Previously: The Designerly Leap Trilogy Part I

Three Pillars in AI and Big Data

Algorithm, modelling, and data are the three foundations behind AI, big data, and analytics in general.

We create a model, write algorithms to describe the internal mechanisms of the model, and then use data to train and evolve the model for it to be practically useful. Sometimes algorithms can help us create new models (2nd generation models).

Intelligence, in business or elsewhere, has become so complex that, we humans can’t do it without the help from intelligent machines. That sounds straightforward, but machine intelligence is not only counterintuitive, but also full of people problem.

Algorithm Modelling and Data.001

Algorithm is Relatively Easy and Openly Abundant

We invent math concepts and techniques well before we find/discover any practical use for them. That has been and will always be the case. We also invent new maths in the progress of attacking difficult problems elsewhere.

The algorithms for AI and Big Data are and will be more abundant. Open source and packaged algorithms are and will continue to be there for everyone to use. Big shot businesses will offer premium algorithms that are customizable, configurable, and ready-to-use.

To anyone who wants algorithms, he/she always has choices.

Both Modelling and Data Have an Inescapable People Problem

Modelling Has a People Problem

Modelling is far more than a scientific challenge.

A model is an abstraction of reality from certain angles. That abstraction is at the sacrifice of real-world context. The abstraction directly reflects what’s relevant to the way the model is to be used.

That relevancy is determined by both scientific evidence and human judgement. The relevancy is tightly coupled with the eventual goals and the measurement we choose to use for them. Those goals and measurement almost always involve how we define success criteria, and how we actually work with the operation of an org.

Therefore that relevancy is unique to each org.

The implication is that, whatever a vendor can commoditize and sell to you, it’s not going to be part of your org’s core value. That core value can only be delivered by the people of your org. They are the stakeholders who are actually accountable for creatively infusing what they know/understand as the core value of the org into the model.

Vendors can and will always offer to help, but they can never be the people who craft the critical part of the model, because they don’t have the deep understanding of and tightly coupled involvement with your org; they can always assist the people of the org on achieving that end, but they’re never the end.

Data Has a Bigger People Problem

Obviously, good or quality data is needed to train, refine, and evolve the model.

The criteria for “good” or “quality” directly reflects what we mean by a good model, how we measure our eventual goals and how we work. That criteria reflects the relevancy in the way we construct the data. That relevancy is, again, tightly coupled with the same higher level measurement and value propositions we have for the case of modelling. The relevancy is, again, unique to each org.

We can’t expect outsiders to create or discover that intimate relevancy for us. We can only expect them to help our own people to do that.

Eventually it’s always the people inside the org that are fundamentally accountable for the core values in that relevancy. Vendors can sell you commoditized patterns, practices, knowledge, and even wisdom; they can never offer unique core values that are supposed to be yours and yours alone.

Algorithm Modelling and Data.002

If you want your org to get ahead or at least cope up with the evolving landscape of big data and AI, you better start from org transformation.

Paving the Way for AI and Big Data in Your Org

Your org rely on your people, and their irreplaceable human judgement, to establish the foundation of intelligent machines that will become part of your org’s ecosystem.

A good start would be evolving work culture and treating your employees well. Without that, your org have no basis for survival in the intelligent-machine-assisting world of the future.

You need to address the people problem to get ahead with the machine problem.

If your employees can’t collaborate with you on a trust basis, you’re merely much less likely to find sustainable ways to work with machines. Nothing wrong with that, you’d simply have a much narrower scope for the kind of businesses you can do, if at all.

It always works from people to machine, not the other way around.

Modelling and data are all about people, because no one but your employees can code/infuse your org’s “DNA” into the model and data that the intelligent machines feed on.

The daunting implication is that, orgs need to become transformable to embrace the future survival with AI and big data, because only the constant transformation can resolve the people problem.

Next: The Designerly Leap Trilogy Part III (Coming soon)

 

{END}

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s